• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

FX-8320 Socket vs Core Temp

There are only 2 temperatures for AMD CPU/APU: socket temp and TCTL. Socket temp is a fair way to judge idle temperature, but is way off under heavy load. Only AOD interprets TCTL correctly by calling it thermal margin, and it is very accurate at load. You can test it yourself. If you get close to zero thermal margin, you will throttle. It works exactly as in the diagram. When people throw these 60'C-70'C temperatures around, you know they are measuring wrong because if you got that high on socket temp you would have already tripped the TCTL cutback due to hitting the thermal margin. If you think I, or AMD engineers are lying, test it yourself. I did.

I don't think anyone is lying I'm just looking for clarity. I've been out of the loop for a while, roughly 3-4 years i'd say.

I say three temps because I have the core, socket and the thermal margin. As far as i know no one knows the max temps for AMD FX ? And if they did would it be the core or socket max temps?

So what your saying is at load I should ignore the CPU temp reported by OCCT and other temp monitors that claim it is hitting 77C and just go off of the Thermal Margin? Also as long as core temp doesn't hit 70C I should be fine but that is inaccurate at load?
 
I don't think anyone is lying I'm just looking for clarity. I've been out of the loop for a while, roughly 3-4 years i'd say.

I say three temps because I have the core, socket and the thermal margin. As far as i know no one knows the max temps for AMD FX ? And if they did would it be the core or socket max temps?

So what your saying is at load I should ignore the CPU temp reported by OCCT and other temp monitors that claim it is hitting 77C and just go off of the Thermal Margin? Also as long as core temp doesn't hit 70C I should be fine but that is inaccurate at load?


It's because AMD got away from using "max temp" and is now using thermal margin. The whole idea of "Core Temp" is now obsolete as there is no actual temperature to measure. It's a calculation, and it's not even on the Celcius scale. But as weird as that sounds, the calculation is very precise. When you approach thermal margin, it will trip and set your clcoks back to a lower power state. I experimented this and it really works, even with all the power saving and thermal throttling safety features disabled in the BIOS it still works.
 
You are not wrong damric and its worth mentioning Aod is now much better and its stability test imho is THE one to use to check max clock stability whilst watching the cores margins.
It can be quite confusing though , with Hwinfo for eg I get three Different cpu core temp readings and there never the same values and are read as degrees C max.
Stick with Aod for T margin temps
 
So as long as Thermal Margin isn't at 0 or in the negatives for prolonged periods of time I can consider myself on the safe side?
 
That's it mate im at 5ghz stable due to this exact realisation.
 
For those of you that still love applications like Core Temp, you can compare side by side to get a reference of what temp reading is getting close to the thermal margin, then you get best of both worlds.
 
A good engineer is always a wee bit conservative, at least on paper ~montgomery scott (scotty)
everyone tends to stress over TEMPS trust us the system will BSOD LONG before you do any real damage you can worried about how hot it is if you have a problem until then stfu about it ;)

example @ 4Ghz @ 1.488V my old phenom II would BSOD precisely at 68.5C if keep below 64C it would run all day at 100% load witch with no AC and 85 to 90F room temp was about where it ran ALL the time
 
A good engineer is always a wee bit conservative, at least on paper ~montgomery scott (scotty)
everyone tends to stress over TEMPS trust us the system will BSOD LONG before you do any real damage you can worried about how hot it is if you have a problem until then stfu about it ;)

example @ 4Ghz @ 1.488V my old phenom II would BSOD precisely at 68.5C if keep below 64C it would run all day at 100% load witch with no AC and 85 to 90F room temp was about where it ran ALL the time
I always had better luck keeping it under 60, once it got over 60 things would perform worse even if the clock was higher. Maybe my chip wasn't very leaky and couldn't tolerate heat as well as yours though.
http://valid.canardpc.com/2627699
 
So again still not sure what to go off of? My idle temps are at 40C on boot it was 32C. That sounding about normal? Playing Watch Dogs it goes to 52C max. OCCT did bring it to 77C though.


Get a temp gun and aim at the base of the heatsink and around the socket...
 
One Caviat on my previous Post is That this comment was in regard to phenom cpu's. I believe this sheds light on amd Fx cpu's.
So I will go back to searching for comment's from the horses mouth so to speak.
 
Sorry to be contrary Buy it is core temp of 61 degrees celsius. I will have some posts if I can find them.
 
it is actually 61.1 degrees celsius fo all FX cpu's except the FX4100 and the FX6100. Their;s is 70 Degree's clelsius core temp.
 
it is actually 61.1 degrees celsius fo all FX cpu's except the FX4100 and the FX6100. Their;s is 70 Degree's clelsius core temp.
What I don't understand is, is that the limit in AMD overdrive is 70C before the thermal margin goes to 0.
 
Be that as it may I just called Amd Technical Support and they also told me My FX9590 was 61.1 core max Temp.
 
Be that as it may I just called Amd Technical Support and they also told me My FX9590 was 61.1 core max Temp.

I'm going to give them a call then too and ask about that.
 
anything below 70c@stock is fine regardless of what AMD says
 
anything below 70c@stock is fine regardless of what AMD says
Aye, i'm running my 8350 at 4,8GHz atm and because I like stuff quiet I've set my fan profile to spin up very late. Core temp regularly peaks at 65C under heavy loads like transcoding, but my chip is rock solid. Could be that the lifespan decreases a bit as a result, but tbh i've yet to hear about someones chip dying from running close to their max temperatures. Its usually voltage that kills a chip.

My former CPU(Fx-8120) didn't really like high temperatures as much though, as it repeatedly crashed at over 60 degrees. Maybe AMD improved their process for their piledriver arch or maybe i'm just lucky/unlucky.

Anyhoo, theres some anecdotal evidence for you :p
 
Back
Top